SPURGEON ET AL . VOL. 8 ’ NO. 1 ’ 894–903 ’ 2014

www.acsnano.org

894

December 08, 2013

C 2013 American Chemical Society

Thickness-Dependent Crossoverfrom Charge- to Strain-MediatedMagnetoelectric Coupling inFerromagnetic/Piezoelectric OxideHeterostructuresSteven R. Spurgeon,† Jennifer D. Sloppy,† Despoina Maria (Demie) Kepaptsoglou,‡

Prasanna V. Balachandran,† Siamak Nejati,§,0 J. Karthik,^ Anoop R. Damodaran,^ Craig L. Johnson, )

Hailemariam Ambaye,# Richard Goyette,# Valeria Lauter,# Quentin M. Ramasse,‡ Juan Carlos Idrobo,4

Kenneth K. S. Lau,§ Samuel E. Lofland, Jr.,3 James M. Rondinelli,† Lane W. Martin,^ and Mitra L. Taheri†,*

†Department of Materials Science and Engineering, Drexel University, Philadelphia, Pennsylvania, United States, ‡SuperSTEM, STFC Daresbury Laboratories,Warrington, United Kingdom, §Department of Chemical and Biological Engineering, Drexel University, Philadelphia, Pennsylvania, United States, ^Department ofMaterials Science and Engineering/Materials Research Laboratory, University of Illinois, Urbana�Champaign, Urbana, Illinois, United States, )Centralized ResearchFacilities, College of Engineering, Drexel University, Philadelphia, Pennsylvania, United States, #Neutron Sciences Directorate, Oak Ridge National Laboratory, OakRidge, Tennessee, United States, 4Center for Nanophase Materials Sciences Division, Oak Ridge National Laboratory, Oak Ridge, Tennessee, United States, and3Department of Physics and Astronomy, Rowan University, Glassboro, New Jersey, United States. 0Present address: Department of Chemical and EnvironmentalEngineering, Yale University, New Haven, Connecticut, United States.

Over the past decade great strideshavebeenmade toward electronicsthat utilize both electron charge

and spin.1,2 For instance, spin-transfer tor-quememories rely on the injection of a spin-polarized current to flip themagnetization ofa free layer in a magnetic tunnel junction.3,4

Direct control of spin polarization wouldgreatly optimize the performance of suchdevices, enabling more robust and efficientcomputing architectures by conveying in-formation through spin transport in the

solid state.5�8 Recent advances in thin-filmgrowth techniques have enabled the syn-thesis of oxide heterostructures wherestrain and charge effects are used to rever-sibly control spin polarization and magne-tization at interfaces.9�15 In particular thereis growing interest in the connection be-tween strain and magnetism in materials,most notably in the active tuning of magne-tization via a coupling of local strain gradi-ents and spin states through the so-called“flexomagnetic” effect.16,17 Flexomagnetism

* Address correspondence tomtaheri@coe.drexel.edu.

Received for review October 29, 2013and accepted December 7, 2013.

Published online10.1021/nn405636c

ABSTRACT Magnetoelectric oxide heterostructures are proposed active layers for

spintronic memory and logic devices, where information is conveyed through spin

transport in the solid state. Incomplete theories of the coupling between local strain,

charge, and magnetic order have limited their deployment into new information and

communication technologies. In this study, we report direct, local measurements of strain-

and charge-mediated magnetization changes in the La0.7Sr0.3MnO3/PbZr0.2Ti0.8O3 system

using spatially resolved characterization techniques in both real and reciprocal space.

Polarized neutron reflectometry reveals a graded magnetization that results from both

local structural distortions and interfacial screening of bound surface charge from the

adjacent ferroelectric. Density functional theory calculations support the experimental observation that strain locally suppresses the magnetization

through a change in the Mn-eg orbital polarization. We suggest that this local coupling and magnetization suppression may be tuned by controlling the

manganite and ferroelectric layer thicknesses, with direct implications for device applications.

KEYWORDS: spintronics . magnetoelectrics . strain engineering . polarized neutron reflectometry .transmission electron microscopy

ARTIC

LE

SPURGEON ET AL . VOL. 8 ’ NO. 1 ’ 894–903 ’ 2014

www.acsnano.org

895

describes the interactions between strain gradientsand local spins; the presence of varying local strainsmay therefore give rise to a sizable flexomagneticcontribution to magnetization.16�18

The current understanding of localized strain andcharge-transfer effects on magnetization is limited,since previous studies have relied on nonlocal probesthat are unable to directly map strain and valencechanges.19 Studies of magnetoelectric heterostruc-tures of the ferromagnetic, half-metal La1�xSrxMnO3

(LSMO) and the piezoelectric PbZrxTi1�xO3 (PZT) ex-emplify the inherent complexity of these systems.Previous work has found that charge-transfer screen-ing of the adjacent ferroelectric layer is largely respon-sible for coupling in ultrathin (<4 nm) LSMO films onPZT,20�22 while other studies have shown that varia-tions in layer thickness and interfacial strain can alsoaffect magnetization.23�28 In these studies the localstrain state of the LSMO/PZT interface was not mea-sured. The relationship between interfacial strain andchemistry is also an important consideration in con-trolling the behavior of these materials, since previousstudies have shown that strain fields around disloca-tions can act as fast paths for interfacial interdiffusionin LSMO/PZT.29,30 It remains unclear how local strainsevolve as a function of layer thickness, how strain andcharge-transfer screening act in concert to mediateinterfacial magnetization, and, more importantly, howto deterministically control this behavior.To better understand flexomagnetism and magne-

toelectric coupling in oxides, it is necessary to movebeyond bulk probes of strain and magnetization to-ward local measurements of strain and interfacialcharge-transfer screening.31�33 Here we synthesizeheterostructures with different local strain and polar-ization states. Using a combination of local atomic andmagnetic characterization, in conjunction with densityfunctional theory (DFT) calculations, we find evidencefor significant strain-induced magnetization changes.We show that large strain changes occur throughoutthe magnetic layer and that they can be tuned by anappropriate choice of substrate thickness. Further-more, we show evidence for interfacial charge-transferscreening, which is secondary to dominant straineffects in thicker layers. Our analysis suggests that itis possible to favor a particular coupling mode by anappropriate choice of ferromagnet and ferroelectriclayer thickness. By using local probes of structure andmagnetization we are able to resolve strain and mag-netization changes within each layer that would beinseparable by bulk techniques.

RESULTS AND DISCUSSION

We used a substrate-induced self-poling techniqueto vary the electrostatic boundary conditions of thebottom electrode interface, so as to pole the PZT awayfrom (on La0.7Sr0.3MnO3 (LSMO)) or toward (on SrRuO3

(SRO)) the substrate, which we term as poled-up and-down, respectively.34�38 Using this method it is pos-sible to control the polarization of the PZT without theneed for large, leaky planar electrodes that wouldpreclude neutron measurements. Four heterostruc-tures were deposited on single-crystal SrTiO3 (001)substrates by pulsed laser deposition (PLD). Oxidemetal underlayers of either LSMO or SRO were depos-ited on a bulk SrTiO3 substrate, followed by either a“thick” (23�37 nm) or “thin” (13 nm) PbZr0.2Ti0.8O3

layer and a cap of ∼10�19 nm LSMO, as shown inFigure 1. These thicknesses were chosen to explore thechanges in strain profiles associated with gradualrelaxation of PZT to the bulk.Aberration-corrected scanning transmission elec-

tron microscopy (STEM) was conducted to confirmthe quality of the LSMO/PZT interfaces. High-angleannular dark field (HAADF) images show that the layerthicknesses are nominally constant in the plane of thefilm (Figure 1 and Supporting Information). The rever-sal of the PZT polarization between the LSMO and SROunderlayers is also confirmed locally by measuring theTi4þ cation displacement at several points along theinterface (Figure 1C,G).39 Since all the film layers weregrown in situ, it was not possible to conduct piezo-response force microscopy (PFM) measurements with-out disturbing the pristine interfaces between eachlayer. X-ray diffraction (XRD) shows that, in-plane, thefilms are constrained to the substrate (see SupportingInformation). However, as we later discuss, the localstrain state of the top LSMO layer varies greatlydepending on the choice of underlayer and PZTthickness.Macroscopic magnetic hysteresis measurements

(Figure 2A,B) reveal a thickness-dependent saturationmagnetization (MS). The data shown have been nor-malized to the entire thickness of LSMOpresent in eachsample. A remarkable 50% (∼0.6 μB/Mn) difference inMS occurs between poled-up and -down heterostruc-tures based on thick PZT (Figure 2A). A smaller 10�20%(0.1�0.2 μB/Mn) difference in MS occurs betweenpoled-up and -down heterostructures based on thinPZT (Figure 2B). For comparison, MS ≈ 1 μB/Mn isexpected for La0.67Sr0.3MnO3 at room temperature.40

These differences are also reflected in the Curie tem-perature (TC) (Figure 2C): the samples deposited on thethin PZT have a TC of 335�342 K, while the samples onthick PZT show a TC of 328�331 K, compared to anominal bulk TC of ∼360 K.41

To probe the local origin of these magnetizationdifferences, polarized neutron reflectometry (PNR) wasconducted at 298 K with an in-plane magnetic field of1 T. Magnetization depth profiles (Figure 2D�G) showthat the MS of the top LSMO layer varies spatiallybut is generally suppressed near the vacuum surfaceas well as at the PZT interface, as has been previouslyobserved.42 Strain-induced distortions of LSMO can

ARTIC

LE

SPURGEON ET AL . VOL. 8 ’ NO. 1 ’ 894–903 ’ 2014

www.acsnano.org

896

suppress TC and consequently room-temperaturemagnetization.43�46 The suppression of TC due to strain-induced distortions in LSMO results from changes in

the Mn�O�Mn bond angles that govern electronhopping between the Mn-eg states responsible fordouble exchange.47�49 Because of the sensitivity of

Figure 2. Top: (A, B) In-plane vibrating sample magnetometry (VSM) measurements conducted at 305 K along the [100]substrate direction, showing a∼50% increase in saturation between the poled-up and poled-down thick PZT samples (A) anda 10�20% increase in saturation for the thin PZT samples (B). (C) Moment versus temperature measurements conducted in a100 Oe magnetic field measured on heating show a significant enhancement of TC with decreasing PZT thickness. Bottom:(D�G) Polarized neutron reflectometry (PNR)magnetization depth profiles measured at 298 K andwith an in-planemagneticfield of 1 T along the [100] substrate direction. The insets show themeasured spin asymmetry (Rþþ � R� �)/(Rþþþ R� �) andthe fits to the data. The vertical dashed lines mark the boundaries between adjacent film layers. The lighter (black) lines are amodel that assumes uniform magnetization throughout each LSMO layer, while the heavier (blue) lines are a model thatallows for gradedmagnetization through the LSMO. The arrows in the inset show regions of improved fitting (as indicated bya smaller chi-squared value). There is a clear suppression of magnetization across the majority of the top LSMO layer in D, aswell as suppression near the vacuum and PZT interfaces in the other samples E�G.

Figure 1. (A, E) Illustrationof the twofilm structures used in this study,with the PZTpolarizationdirection indicatedby the arrows.Characteristic high-angle annular dark field (STEM-HAADF) images of the top (B, F) and bottom (D, H) PZT interfaces, showing theabsence of any extrinsic defects. (C, G) Cross-correlated images of the PZT layer, confirming the change in polarization; the insetsare the result of multislice simulations, with the horizontal dash corresponding to the center of the unit cell.

ARTIC

LE

SPURGEON ET AL . VOL. 8 ’ NO. 1 ’ 894–903 ’ 2014

www.acsnano.org

897

the double-exchange mechanism to strain, local strainfluctuations;if present;may give rise to the gradedmagnetization profiles observed in PNR.Geometric phase analysis (GPA) was used to test this

hypothesis by measuring strains directly from TEMmicrographs with ∼0.1% accuracy down to the nano-meter scale.50,51 In all samples the in-plane strain isessentially uniform over the 3�5 nm integration win-dow, varying by <0.1%. We note that there is goodagreement between the average GPA-measured c/aaxial ratios and those measured by XRD (see supple-mental Table SII). The out-of-plane strain relative tobulk unstrained LSMO increases normal to the PZTinterface, reaching a maximum at the vacuum surface.Using this technique we are able to map the local c/aaxial ratio within each sample (Figure 3B). This analysisreveals that for both thick PZT samples (Figure 3C,D)the c/a of the LSMO increases from ∼0.96 at the PZTinterface to 1.01�1.03 at the vacuum surface. Thiscorresponds to a strain gradient of approximately(2.95�4.76) � 106 m�1. The poled-down thin PZTsample (Figure 3F) shows a similar trend, increasingfrom∼0.98 at the PZT interface to∼1.03 at the vacuumsurface (a gradient of 4.76 � 106 m�1). However, thepoled-up thin PZT sample (Figure 3E) shows aU-shaped profile that drops from ∼1.04 at the PZTinterface to ∼0.97 at the middle of the LSMO andincreases to∼1.05 at the vacuum surface. The changescoincide with significantly different c/a ratios in theadjacent PZT layer, which ranges from 1.04 to 1.1,suggesting that the interfacial strain state is heavilydependent on the tetragonality of the underlying PZTlayer, as well as the thickness of the LSMO layer. Moreimportantly, a comparison of the PNR and GPA data

shows that, in general, an LSMO c/a that deviatesoutside of the range 0.98�0.995 coincides with localsuppression of magnetization, which agrees well withchanges in bulk properties.52 The observed strainfluctuations may correlate to local spin changes, parti-cularly since they are comparable in magnitude to thestrains needed to induce a measurable flexoelectriceffect in other systems.53,54 While direct flexomagnet-ism is limited to a subset of symmetry classes, indirectflexomagnetism is expected to be present in all mag-netoelectrics, wherever polarization and magnetiza-tion are coupled.55

To estimate the strain-induced suppression of mag-netization in the samples, we turn to the empiricalmodel of Millis et al.47 and density functional theory(DFT) calculations. Millis et al. proposed a model thatrelates TC to the substrate strain-induced enhance-ment of the Jahn�Teller distortion relative to un-strained bulk LSMO (see Experimental Section).47,56

We choose this model since it allows us to directlysubstitute the averaged local Æc/aæ, extracted from theexperimental GPA, to obtain an estimate of TC. We thenconducted DFT calculations to explore the electron�lattice effects mediating the microscopic coupling indetail.For the poled-up PZT samples we find fromGPA that

Æc/aæLSMO ≈ 0.99�1.01, and we estimate TC ≈ 249�295 K for the top LSMO layer using the Millis et al.model. These out-of-plane strains appear to greatlysuppress the ferromagnetic ordering of the top layer,as is observed in PNR (Figure 2D,F). In contrast, forthe poled-down samples, we find that Æc/aæLSMO ≈0.98�0.995, and we estimate TC ≈ 319�327 K. Thesedistortions result in a higher TC and larger average

Figure 3. (A) Characteristic STEM-HAADFmicrographof the LSMO/PZT interface; the inset shows the fast Fourier transformofthe PZT layer. (B) Characteristicmap of local c/a axial ratios in the LSMO and PZT layers. This ratio varies throughout the LSMObut is largest at the vacuum interface. (C�F) Line scans of c/a normal to the LSMO/PZT interface for all four films. The verticalline indicates the PZT boundary, while the horizontal dashed region indicates the c/a range outside of whichmagnetization isexpected to be suppressed.

ARTIC

LE

SPURGEON ET AL . VOL. 8 ’ NO. 1 ’ 894–903 ’ 2014

www.acsnano.org

898

magnetization across the LSMO (Figure 2E,G). Thepredicted and measured Curie temperatures for thepoled-down samples are in excellent agreement(Figure 2C); however, the agreement for the poled-upsamples is worse, perhaps because these samplesinclude two LSMO layers and the measurement of TCis less accurate.We next perform spin-polarized DFT calculations

within the generalized-gradient approximation plusHubbard-U method on a series of LSMO structures toisolate the contributions of epitaxial strain from inter-facial charge transfer on TC and Mn-eg orbital polariza-tion. We choose lattice constants consistent with theexperimental epitaxial constraints and Æc/aæLSMO ratiosranging from 0.985 (poled-up thick PZT) to 0.994(poled-down thick PZT). We note that these axial ratiosrefer to themetric shape of the simulation cell, not localoctahedral elongations, and thus deviations from thelocal strain measurements determined using GPA areexpected. The atomic positions are then fully relaxed,allowing for rotations and bond elongations. Firstwe computed the optimal Æc/aæ for LSMO on (001)-oriented STO and obtained a value of 0.985, which isconsistent with the average Æc/aæ of LSMO on poled-upthick PZT. We then applied 0.5% and 1% uniaxial strainalong the [001] direction to simulate the range ofobserved axial ratios (Figure 4), as a means to disen-tangle the strain contributions from interface effectsdue to coherent strain of LSMO with varying PZTthickness and polarization. We then calculated amean-field theoretical ferromagnetic Curie tempera-ture, TC

MFT, following the procedure in Kübler et al.57

and Lampis et al.58

Our DFT results indicate that out-of-plane stretchingmonotonically increases TC

MFT from 292 K (0%) to 323 K(1%) at the highest Æc/aæ state. Our TCMFT trend com-pares favorably with our measured poled-up PZT

samples that have LSMO underlayers as well as themodel calculations followingMillis et al. To quantify theorbital occupancy, we calculated the electron orbitalpolarization, P = (nx2�y2 � nz2)/nx2�y2 þ nz2), of Mn-egorbitals from the partial density of states (PDOS)spectra, where nx2�y2 and nz2 are the area under thecurve for dx2�y2 and dz2 orbitals, respectively, integratedup to the Fermi level.59 A positive value forP indicates that electrons favor dx2�y2 orbital occu-pancy, and a negative P value indicates that electronsfavor dz2 occupancy. We find that in bulk unstrainedLSMO P takes both negative and positive values, andthe magnitude of P is roughly the same for both Mnsites (Figure 4 and Figure S9); TC

MFT for bulk LSMO isestimated to be 388 K. Application of an in-planetensile strain alone promotes preferential dx2�y2 orbitalfilling; P takes only positive values at bothMn sites, andTCMFT reduces drastically to 292 K, which agrees wellwith our experimental measurements made on poled-up samples. Uniaxial strain along [001] gradually trans-fers charge to the Mn dz2 orbital aligned along thez-direction, as expected.44,60,61 At 1% elongation,P becomes both negative and positive (TC

MFT ≈ 322 K),albeit reduced relative to unstrained LSMO. Commen-surate with the filling of dz2 orbitals as a function of out-of-plane stretching, TC

MFT is also found to increase,indicating a direct association between the c/a axialratio, macroscopic TC

MFT, and P in LSMO.Although a clear trend emerges between c/a, TC, and

P, in agreement with previous literature,62,63 we recog-nize that our DFT calculations do not fully capture theTC behavior of the poled-up samples. While thesesamples have the largest average Æc/aæ value (as mea-sured by XRD), their measured TC is lower than that ofother samples, indicating the existence of an additionalcompeting mechanism not captured in our simula-tions. We also note that these calculations are done

Figure 4. (A) Relationship between TCMFT (K) and P (in %) for various simulation cells as calculated fromDFT. Positive value for

P indicates the percentage excess of Mn-eg electrons filling the dx2�y2 orbital relative to the dz2 orbital and vice versa. The30-atom supercell contains two distinct Mn atoms, Mn(I) (open, red) and Mn(II) (filled, blue) (see Supporting Information). (B)Relationship between TC

MFT (K) and axial ratio (c/a) as calculated from DFT. A clear trend emerges between c/a, P, and TCMFT. In

unstrained LSMO, both dx2�y2 and dz2 are filled. The application of in-plane tensile strain promotes preferential dx2�y2 filling inboth Mn atoms; simultaneously TC

MFT decreases. However, out-of-plane stretching gradually promotes transfer of charge todz2 orbitals, and a corresponding gradual increase in TC

MFT is found. Circles correspond to bulk LSMO, and triangles areepitaxially strained LSMO (under uniaxial strain varying from 0 to 1% along the [001] direction).

ARTIC

LE

SPURGEON ET AL . VOL. 8 ’ NO. 1 ’ 894–903 ’ 2014

www.acsnano.org

899

assuming a uniform uniaxial strain, in contrast to thechanging strain observed in GPA; nonetheless, webelieve that these results provide a valuable insightinto how increasing tetragonality affects electronic andmagnetic ordering.To probe other possible coupling mechanisms,

such as chemistry changes or charge-transfer screen-ing, we have conducted electron energy loss spectros-copy (EELS) mapping of the LSMO/PZT interface. TheMn L23 white lines near 640�665 eV are measured inthis study since they contain information about excita-tions from the spin�orbit split 2p3/2 and 2p1/2 levels toavailable states in the 3d band.64�69 Screening ofsurface charge from the adjacent PZT layer gives riseto a change in the local 3d band occupancy, reflectedin a deviation from the nominal Mn3þ/Mn4þ ratio of∼3.3.70,71

Figure 5 shows the results of STEM-EELS maps at theLSMO/PZT interface for the thick samples. We find thatthe interfaces are quite sharp, with the EELS signallimited to one to two atomic planes away from theinterface; however, because STEM is a localized tech-nique, it is impossible to completely rule out someintermixing in either interface. Both samples possess abulk valence of ∼3.4 (near the nominal 3.3), but at theinterface the value for the poled-up sample drops to∼2.63, while that of the poled-down sample increasesto∼4.26. Additionally, there is a clear shift of the Mn L3edge toward lower energy in the poled-up sample

(Figure 5B), indicating lower valence; however, the shiftin the poled-down sample is not as pronounced.68,70

The valence change is spread over 3�4 unit cells at theinterface, with an average valence at ∼3.02 for poled-down and∼3.89 for poled-up, a difference of∼0.87. Itshould be noted that the error bars on this data are stillrather large, ruling out more detailed analysis of theinduced valence, but there is clearly an interfacialchange likely resulting from interaction with the ad-jacent ferroelectric layer.Thus, while local strain fluctuations suppress mag-

netization across larger length scales, it appears thatcharge-transfer screening operates in a ∼2 nm inter-face region at the PZT boundary (Figure 2D�G), in linewith prior estimates.72�74 PNR measurements showthat the change between states in the thick samples at298 K is Δm =mdown �mup = 0.88�0.22 = 0.66 μB/Mn,while, for the thin samples,Δm=0.54�0.34=0.20μB/Mn.These values agree well with previous magneto-opticalstudies of ultrathin LSMO that foundΔm = 0.76 μB/Mn.20

CONCLUSIONS

Several trends are now clear. We find that it ispossible to self-pole PZT through the use of an appro-priate substrate material, a method that may be ex-tended tomany other systems. Macroscopic bulkmag-netizationmeasurements show thatMS and TC dependon both PZT polarization and thickness. PNR measure-ments reveal thatMS varies locally and ismost suppressed

Figure 5. High-angle annular dark field images and electron energy loss spectroscopymaps of the top LSMO/PZT interface inthe poled-up (A) and poled-down (D) thick PZT samples. The numbers indicate the atomic rows across which average spectrawere collected and correspond to the Mn L23 spectra in B and E. (C, F) Calculated Mn L3/L2 ratios and estimated Mn valencesfrom each row. Error bars correspond to the standard error of the Gaussian fits to the edges. Although both samples possessthe same valence in the bulk (∼3.4), they diverge near the PZT interface, indicating screening of surface charge from theadjacent PZT layer.

ARTIC

LE

SPURGEON ET AL . VOL. 8 ’ NO. 1 ’ 894–903 ’ 2014

www.acsnano.org

900

at the LSMO/vacuum interface, where the GPA-measured LSMO c/a axial ratio is largest. Furthermore,wefindevidence for large strain gradients (∼106m�1) inthe LSMO. Phenomenological models show that localstrains affect the Mn-eg electronic distribution andplay a role in suppressing the LSMO TC. EELS and PNRalso reveal the presence of a ∼2 nm charge-transferscreening region that affects magnetization at thePZT interface. However, the magnitude of the inducedmagnetization does not directly agree with the pre-vious work of Vaz et al., suggesting that other factorsmay be at work.20

Collectively the results obtained in this study sug-gest a more complex model of strain- and charge-mediated magnetization in ferroelectric/ferromagnetcomposites. We find that the tetragonality of the PZThas a pronounced effect on the interfacial strain in theLSMO: a larger PZT c/a corresponds to a larger inter-facial c/a in the LSMO, which gradually increases nearthe vacuum surface. In the ultrathin limit (<4 nm), strain

fluctuations in LSMO are minimal and charge-transferscreening drives coupling. As the LSMO thicknessincreases, local strain fluctuations soon overwhelmthe magnetization of the layer, indicating that layergeometries are crucial components in the design ofthese materials. In excess of the ultrathin limit, ourPNR results indicate that local strain can inducemuch larger changes in the magnetization profileof LSMO than charge-transfer screening. By tuningthe PZT tetragonality through doping or an appro-priate substrate, it is possible to reshape magnetiza-tion gradients in the ferromagnet. Our resultssuggest that a piezoelectric substrate may be usedto actively control local strain and directly vary thespin state of the ferromagnet. The wealth of insightprovided by this suite of techniques shows that localprobes of magnetization, strain, and chemistry arean invaluable way to understand coupling of multi-ple degrees of freedom in magnetoelectrics andemerging flexomagnets.

EXPERIMENTAL SECTIONSrRuO3 and PbZr0.2Ti0.8O3 films were grown by PLD at 635 �C

at 100 and 200 mTorr pO2, with laser repetition rates of 12 and3 Hz and laser fluences of 1.75 and 2 J cm�2, respectively.The La0.7Sr0.3MnO3 layers were grown at 650 �C at 200 mTorrof oxygen with a laser repetition rate of 2 Hz and a fluence of1.5 J cm�2. Films were then cooled to room temperature in760 Torr pO2.The crystallinity of the as-grown films was measured by XRD

with Cu KR radiation (λ = 0.15418 nm) on a Panalytical Empyreandiffractometer. Reciprocal spacemapsweremade around the STO103 diffraction condition. Layer thickness was studied by X-rayreflectivity as measured on a Rigaku SmartLab diffractometer.Bulk magnetometry was conducted with a Quantum Design

vibrating sample magnetometer at 305 K along the [100] and[110] in-plane substrate directions, with no discernible differ-ence in hysteresis. TC was measured in the range 310�350 Kunder an applied in-plane magnetic field of 100 Oe. AnArrott�Belov analysis was conducted to determine TC, assum-ing self-consistent samples (see Supporting Information).PNR was conducted at 298 K with an in-plane magnetic field

of 1 T applied along the [100] substrate direction. Non-spin-flip specular reflectivites were measured from q = 0.005 to0.1 �1. The reflectivity data were then fit with the ReflPaksoftware package and refined in conjunction with XRD.A fit was conducted with uniform magnetization in theLSMO layers, and a second fit was conducted in which themagnetization was allowed to vary. The latter resulted in abetter fit to the measured spin asymmetry, particularly athigher q.Samples were prepared for TEM by conventional mechanical

polishing and ion milling. HRTEM images were captured at200 keV on a JEOL 2100 LaB6. BF-STEM and STEM-HAADFmicrographs were also captured on a CS-corrected FEI TitanSTEM operating at 300 keV. EELSmaps and HAADF images weremeasured on a CS-corrected Nion UltraSTEM 100 operating at100 keV, with a convergence angle of 30 mrad and an effectiveenergy resolution of 0.6�0.7 eV.75 The background was re-moved from each scan using a power law fit, and spectra wereextracted from eachmap row-by-rowwith a∼0.1� 0.8� 1 nm2

window. Hartree-Slater cross sections were subtracted fromeach edge, and the spectra were processed with the EELSToolspackage in Digital Micrograph to extract Mn L23 ratios from thepositive component of the second derivative.76

Cation displacements were determined from a series of10�40 STEM-HAADF acquisitions, which were captured at 5 μsintervals and cross-correlated and averaged with the ImageJprogram with the StackReg plugin. Image simulations along thePZT [100] direction were conducted with the multislice methodin the QSTEM program.77 This allows us to achieve a precision tomeasure the atomic displacements better than∼8 pm.78 Severalatomic displacements were modeled with a 69 � 70 � 160supercell consisting of 80 slices. A 400 � 400 pixel array with a0.05Å� 0.05Å resolution and 20Å� 20Åwindow sizewas used,along with the microscope parameters.GPA was conducted on STEM-HAADF and HRTEM images

displaying minimal drift or scan error. First maps of local re-ciprocal lattice vectors corresponding to out-of-plane (g1)and in-plane (g2) directions were constructed. The ratio ofthese two maps (g2/g1) then gives the local c/a.79 The lineprofiles shown in Figure 4 were measured by integrating3�5 nm in-plane to minimize noise. It should be noted thatlocal contrast and thickness fluctuations can give rise to localspikes in the measured ratio, so we only discuss broadertrends in c/a.To calculate in- (εxx) and out-of-plane (εyy) LSMO strains,

references were chosen in either the STO or PZT layers; in thelatter the measured strain values have been shifted to accountfor the average strain across the PZT layer. Themeasured strainsin the top LSMO layer were converted relative to bulk LSMOaccording to

εrelative ¼ εmeasured þ cSTO,bulk� cLSMO,bulkcSTO,bulk

!cSTO,bulkcLSMO,bulk

where cSTO,bulk = 3.905 Å and cLSMO,bulk = 3.87 Å.TC was estimated from these GPA strains using the empirical

model of Millis et al.47

TC(ε) ¼ TC(ε ¼ 0) 1 � RεB� 12ΔεJT

2

� �

where εB = (2εxxþ εyy) and εJT = (2/3)1/2(εyy� εxx). TC(ε = 0) is thebulk LSMO TC of∼360 K, while R and Δ are empirical constantsthat represent the weighting of the bulk strain and Jahn�Tellerdistortion of MnO6 octahedra, respectively. Typical values areR ≈ 10 and Δ ≈ 270.56

DFT calculations were performed within the spin-polarizedgeneralized gradient approximation (GGA) plus Hubbard-U

ARTIC

LE

SPURGEON ET AL . VOL. 8 ’ NO. 1 ’ 894–903 ’ 2014

www.acsnano.org

901

method as implemented in the Quantum-ESPRESSO packageversion 5.0.80 The Dudarev et al. approach80 was followed toinclude an effective Hubbard term of 3 eV for unstrained LSMOand 2 eV for LSMO on STO to accurately treat the correlated Mn3d electrons. The core and valence electrons were treated withthe ultrasoft pseudopotential81 and the PBEsol exchange�correlation functional.82,83 The Brillouin-zone integrations wereperformed with a Marzari�Vanderbilt smearing84 of 0.02 Ryover a 7 � 7 � 5 Monkhorst�Pack k-point mesh85 centered atΓ and a 60 and 600 Ry plane-wave and kinetic energy cutoff forcharge density, respectively. For density of states (DOS) calcula-tions, a denser 14 � 14 � 12 Monkhorst�Pack k-point meshsampling was used. Atomic positions were allowed to convergeuntil the Hellmann�Feynman forces became less than 2 meV Å1.Structure optimization was performed using the Broyden�Fletcher�Goldfarb�Shanno (BFGS) algorithm. The approachadapted byMa et al.86 was used to simulate the crystal structureof LSMO (see Supporting Information). Orbital polarization (P)was calculated using the formula

P ¼ nx2� y2 � nz2

nx2� y2 þ nz2

where nx2�y2 and nz2 are the area under the partial density ofstates spectra of dx2�y2 and dz2 orbitals respectively (for bothspins) within the energy window from the Fermi level to �8 eVbelow it. The nearest-neighbor exchange coupling constant J0was calculated within the mean-field approximation57,58 with

J0 ¼ EF� EAFM� A

12(∑SF1S

F2 �∑SAFM� A

1 SAFM � A2 )

where EF and EAFM�A are the total energies (eV) of spin-polarizedferromagnetically ordered and spin polarized A-type antiferro-magnetically ordered calculations, respectively, S1 is the calcu-lated atomic magnetic moment of the Mn(I) atom, and S2 is thecalculated atomic magnetic moment of the Mn(II) atom. FromJ0, T̂C was estimated by the mean-field theory approximation,T̂C = (2/3)S(S þ 1)(J0/kB),

57,58 where S = ((4S1F) þ (2S2

F))/6 is theweighted average of the magnetic moments of Mn in theferromagnetic spin order configuration.

Conflict of Interest: The authors declare no competingfinancial interest.

Acknowledgment. S.R.S. and M.L.T. thank Drs. Steven Mayand Rebecca Sichel-Tissot for constructive discussions. S.R.S.also thanks Christopher R. Winkler and Michael L. Jablonskifor their assistance with TEM sample preparation. The authorsgratefully acknowledge support from the National Science Foun-dation under grants CMMI-1031403 (M.L.T.), DMR-0908779 (S.E.L),DMR-0821406 (S.E.L.), CBET-0846245 (S.N. and K.K.S.L), andDMR-1149062 (J.K. and L.W.M.), as well as from the Office ofNaval Research under grants N00014-1101-0296 (M.L.T.) andN00014-10-1-0525 (J.K. and L.W.M). A.R.D. and L.W.M. acknowl-edge support from the Army Research Office under grantW911NF-10-1-0482. P.V.B. and J.M.R. were supported by theDefense Advanced Research Projects Agency under grantN66001-12-4224. Electron microscopy and X-ray photoelectronspectroscopy (NSF MRI CBET-0959361) were conducted inDrexel University's Centralized Research Facilities, and X-rayreciprocal space mapping and magnetic measurements wereperformed in Rowan University's Department of Physics andAstronomy. X-ray diffraction work was also conducted in theLaboratory for Research on the Structure of Matter at theUniversity of Pennsylvania. DFT modeling was conducted onthe Cray XE6 Garnet system at the U.S. Army Engineer Researchand Development Center. Part of this research was a userproject supported by Oak Ridge National Laboratory's Centerfor Nanophase Materials Sciences (CNMS), which is sponsoredby the Scientific User Facilities Division, Office of Basic EnergySciences, U.S. Department of Energy (J.C.I.). Electronmicroscopywas carried out in part at SuperSTEM, the U.K. National Facilityfor Aberration-Corrected STEM supported by the U.K. Engineer-ing and Physical Sciences Research Council. Polarized neutronreflectometry experiments were performed at the Spallation

Neutron Source at Oak Ridge National Laboratory, managed byUT-Battelle, LLC, for the U.S. Department of Energy. Author S.R.S.is supported by a National Science Foundation IntegrativeGraduate Education and Research Traineeship (IGERT) and aDepartment of Defense National Defense Science and Engi-neering Graduate (NDSEG) Fellowship.

Supporting Information Available: Structural characteriza-tion and experimental details are available free of charge viathe Internet at http://pubs.acs.org/.

REFERENCES AND NOTES1. Bader, S.; Parkin, S. Spintronics. Annu. Rev. Condens. Matter

Phys. 2010, 1, 71–88.2. Fert, A. The Present and the Future of Spintronics. Thin

Solid Films 2008, 517, 2–5.3. Stiles, M.; Zangwill, A. Anatomy of Spin-Transfer Torque.

Phys. Rev. B 2002, 66, 014407.4. Sankey, J. C.; Cui, Y.-T.; Sun, J. Z.; Slonczewski, J. C.; Buhrman,

R. A.; Ralph, D. C. Measurement of the Spin-Transfer-TorqueVector in Magnetic Tunnel Junctions. Nat. Phys. 2007, 4,67–71.

5. Prinz, G. A. Magnetoelectronics. Science 1998, 282, 1660–1663.

6. Wolf, S.; Treger, D. Spintronics: A New Paradigm for Elec-tronics for the New Millennium. IEEE Trans. Magn. 2000,36, 2748–2751.

7. Fert, A.; George, J.-M.; Jaffrès, H.; Mattana, R.; Seneor, P. TheNew Era of Spintronics. Europhys. News 2003, 34, 227–229.

8. Wolf, S. A.; Awschalom, D. D.; Buhrman, R. A.; Daughton,J. M.; von Molnár, S.; Roukes, M. L.; Chtchelkanova, A. Y.;Treger, D. M. Spintronics: A Spin-Based Electronics Visionfor the Future. Science 2001, 294, 1488–1495.

9. Eerenstein, W.; Wiora, M.; Prieto, J. L.; Scott, J. F.; Mathur,N. D. Giant Sharp and Persistent ConverseMagnetoelectricEffects in Multiferroic Epitaxial Heterostructures. Nat.Mater. 2007, 6, 348–351.

10. Garcia, V.; Bibes, M.; Bocher, L.; Valencia, S.; Kronast, F.;Crassous, A.; Moya, X.; Enouz-Vedrenne, S.; Gloter, A.;Imhoff, D.; et al. Ferroelectric Control of Spin Polarization.Science 2010, 327, 1106–1110.

11. Thiele, C.; Dörr, K.; Schultz, L.; Beyreuther, E.; Lin, W.-M.Piezoelectrically Induced Resistance Modulations inLa0.7Sr0.3MnO3/Pb(Zr,Ti)O3 Field Effect Devices. Appl. Phys.Lett. 2005, 87, 162512.

12. Martin, L.; Chu, Y.-H.; Ramesh, R. Advances in the Growthand Characterization of Magnetic, Ferroelectric, and Multi-ferroicOxide Thin Films.Mater. Sci. Eng., R 2010, 68, 89–133.

13. Wang, Y.; Hu, J.; Lin, Y.; Nan, C.-W. Multiferroic Magneto-electric Composite Nanostructures. NPG Asia Mater. 2010,2, 61–68.

14. Bibes, M.; Barthélémy, A. Multiferroics: Towards a Magne-toelectric Memory. Nat. Mater. 2008, 7, 425–426.

15. Béa, H.; Gajek,M.; Bibes,M.; Barthélémy, A. SpintronicswithMultiferroics. J. Phys.: Condens. Matter 2008, 20, 434221.

16. Lukashev, P.; Sabirianov, R. F. Flexomagnetic Effect inFrustrated Triangular Magnetic Structures. Phys. Rev. B2010, 82, 094417.

17. Eliseev, E.; Morozovska, A.; Glinchuk, M.; Blinc, R. Sponta-neous Flexoelectric/Flexomagnetic Effect in Nanoferroics.Phys. Rev. B 2009, 79, 165433.

18. Sabirianov, R. F.; Lukashev, P. V.Magneto-Elastic Propertiesof Frustrated Triangular Magnetic Structure: FlexomagneticEffect. 2010; http://meetings.aps.org/link/BAPS.2010.MAR.Q34.8.

19. Vaz, C. A. F. Electric Field Control of Magnetism in Multi-ferroic Heterostructures. J. Phys.: Condens.Matter 2012, 24,333201.

20. Vaz, C. A. F.; Hoffman, J.; Segal, Y.; Reiner, J. W.; Grober, R. D.;Zhang, Z.; Ahn, C. H.; Walker, F. J. Origin of the Magneto-electric Coupling Effect in Pb(Zr0.2 Ti0.8)O3/La0.8 Sr0.2 MnO3

Multiferroic Heterostructures. Phys. Rev. Lett. 2010, 104,127202.

21. Molegraaf, H. J. A.; Hoffman, J.; Vaz, C. A. F.; Gariglio, S.;van der Marel, D.; Ahn, C. H.; Triscone, J.-M. Magnetoelectric

ARTIC

LE

SPURGEON ET AL . VOL. 8 ’ NO. 1 ’ 894–903 ’ 2014

www.acsnano.org

902

Effects in Complex Oxides with Competing Ground States.Adv. Mater. 2009, 21, 3470–3474.

22. Vaz, C. A. F.; Segal, Y.; Hoffman, J.; Grober, R. D.; Walker, F. J.;Ahn, C. H. Temperature Dependence of the Magneto-electric Effect in Pb(Zr0.2 Ti0.8)O3/La0.8Sr0.2MnO3 Multifer-roic Heterostructures. Appl. Phys. Lett. 2010, 97, 042506.

23. Dörr, K.; Thiele, C. Multiferroic Bilayers of Manganites andTitanates. Phys. Status Solidi B 2006, 243, 21–28.

24. Hezareh, T.; Razavi, F. S.; Kremer, R. K.; Habermeier, H.-U.;Lebedev, O. I.; Kirilenko, D.; Tendeloo, G. V. Effect ofPbZr0.52Ti0.48O3 Thin Layer on Structure, Electronic and Mag-netic Properties of La0.65Sr0.35MnO3 and La0.65Ca0.30MnO3

Thin-Films. J. Appl. Phys. 2011, 109, 113707.25. Zheng, R.; Wang, Y.; Liu, Y.; Gao, G.; Fei, L.; Jiang, Y.; Chan, H.;

Li, X.; Luo, H.; Li, X. Epitaxial Growth and Interface StrainCoupling Effects in Manganite Film/Piezoelectric-CrystalMultiferroic Heterostructures. Mater. Chem. Phys. 2012,133, 42–46.

26. Vrejoiu, I.; Ziese, M.; Setzer, A.; Esquinazi, P. D.; Birajdar, B. I.;Lotnyk, A.; Alexe, M.; Hesse, D. Interfacial Strain Effects inEpitaxial Multiferroic Heterostructures of PbZrxTi1‑xO3/La0.7Sr0.3MnO3 Grown by Pulsed-Laser Deposition. Appl.Phys. Lett. 2008, 92, 152506.

27. Shu, L.; Li, Z.; Ma, J.; Gao, Y.; Gu, L.; Shen, Y.; Lin, Y.; Nan, C.W.Thickness-Dependent Voltage-Modulated Magnetism inMultiferroic Heterostructures. Appl. Phys. Lett. 2012, 100,022405.

28. Liu, M.; Li, S.; Obi, O.; Lou, J.; Rand, S.; Sun, N. X. Electric FieldModulation of Magnetoresistance in Multiferroic Hetero-structures for Ultralow Power Electronics. Appl. Phys. Lett.2011, 98, 222509.

29. Arredondo, M.; Ramasse, Q. M.; Weyland, M.; Mahjoub, R.;Vrejoiu, I.; Hesse, D.; Browning, N. D.; Alexe, M.; Munroe, P.;Nagarajan, V. Direct Evidence for Cation Non-Stoichiometryand Cottrell Atmospheres aroundDislocation Cores in Func-tional Oxide Interfaces. Adv. Mater. 2010, 22, 2430–2434.

30. Arredondo, M.; Weyland, M.; Hambe, M.; Ramasse, Q. M.;Munroe, P.; Nagarajan, V. Chemistry of Ruddlesden-Popper Planar Faults at a Ferroelectric-FerromagnetPerovskite Interface. J. Appl. Phys. 2011, 109, 084101.

31. Chakhalian, J.; Millis, A. J.; Rondinelli, J. Whither the OxideInterface. Nat. Mater. 2012, 11, 92–94.

32. Wolf, S. A.; Lu, J.; Stan, M. R.; Chen, E.; Treger, D. M. ThePromise of Nanomagnetics and Spintronics for FutureLogic and Universal Memory. Proc. IEEE 2010, 98, 2155–2168.

33. Raveau, B.; Maignan, A.; Mahendiran, R.; Khomskii, D.;Martin, C.; Hébert, S.; Hervieu, M.; Frésard, R. Instability ofMagnetism and Conductivity in CMR Manganites: Role ofMn-Site Doping and Thermal Cycling. J. Phys. Chem. Solids2002, 63, 901–905.

34. Chen, J.; Lu, H.; Liu, H.-J.; Chu, Y.-H.; Dunn, S.; Ostrikov, K.;Gruverman, A.; Valanoor, N. Interface Control of SurfacePhotochemical Reactivity in Ultrathin Epitaxial Ferroelec-tric Films. Appl. Phys. Lett. 2013, 102, 182904.

35. Karthik, J.; Damodaran, A. R.; Martin, L. W. Epitaxial Ferro-electric Heterostructures Fabricated by Selective AreaEpitaxy of SrRuO3 Using an MgO Mask. Adv. Mater. 2012,24, 1610–1615.

36. Yu, P.; Luo, W.; Yi, D.; Zhang, J. X.; Rossell, M. D.; Yang, C.-H.;You, L.; Singh-Bhalla, G.; Yang, S. Y.; He, Q.; et al. InterfaceControl of Bulk Ferroelectric Polarization. Proc. Natl. Acad.Sci. U.S.A. 2012, 109, 9710–9715.

37. Afanasjev, V. P.; Petrov, A. A.; Pronin, I. P.; Tarakanov, E. A.;Kaptelov, E. J.; Graul, J. Polarization and Self-Polarization inThin PbZr1�xTixO3 (PZT) Films. J. Phys.: Condens. Matter2001, 13, 8755–8763.

38. Jia, C.-L.; Nagarajan, V.; He, J.-Q.; Houben, L.; Zhao, T.;Ramesh, R.; Urban, K.; Waser, R. Unit-Cell Scale Mappingof Ferroelectricity and Tetragonality in Epitaxial UltrathinFerroelectric Films. Nat. Mater. 2007, 6, 64–69.

39. Gao, P.; Nelson, C. T.; Jokisaari, J. R.; Baek, S.-H.; Bark, C. W.;Zhang, Y.; Wang, E.; Schlom, D. G.; Eom, C.-B.; Pan, X.Revealing the Role of Defects in Ferroelectric Switchingwith Atomic Resolution. Nat. Commun. 2011, 2, 591.

40. Lu, H.; George, T. A.; Wang, Y.; Ketsman, I.; Burton, J. D.; Bark,C.-W.; Ryu, S.; Kim, D. J.; Wang, J.; Binek, C.; et al. ElectricModulation of Magnetization at the BaTiO3/La0.67Sr0.33MnO3

Interfaces. Appl. Phys. Lett. 2012, 100, 232904.41. Chen, H.; Ismail-Beigi, S. Ferroelectric Control of Magneti-

zation in La1�xSrxMnO3 Manganites: A First-PrinciplesStudy. Phys. Rev. B 2012, 86, 1–13.

42. Freeland, J. W.; Kavich, J. J.; Gray, K. E.; Ozyuzer, L.; Zheng,H.; Mitchell, J. F.; Warusawithana, M. P.; Ryan, P.; Zhai, X.;Kodama, R. H.; et al. Suppressed Magnetization at theSurfaces and Interfaces of Ferromagnetic Metallic Manga-nites. J. Phys.: Condens. Matter 2007, 19, 315210.

43. Adamo, C.; Ke, X.; Wang, H. Q.; Xin, H. L.; Heeg, T.; Hawley,M. E.; Zander, W.; Schubert, J.; Schiffer, P.; Muller, D. A.; et al.Effect of Biaxial Strain on the Electrical and MagneticProperties of (001) La0.7Sr0.3MnO3 Thin Films. Appl. Phys.Lett. 2009, 95, 112504.

44. Angeloni, M.; Balestrino, G.; Boggio, N. G.; Medaglia, P. G.;Orgiani, P.; Tebano, A. Suppression of the Metal-InsulatorTransition Temperature in Thin La0.7Sr0.3MnO3 Films. J. Appl.Phys. 2004, 96, 6387.

45. Tsui, F.; Smoak, M. C.; Nath, T. K.; Eom, C. B. Strain-Dependent Magnetic Phase Diagram of EpitaxialLa0.67Sr0.33MnO3 Thin Films. Appl. Phys. Lett. 2000, 76,2421.

46. Ranno, L.; Llobet, A.; Tiron, R.; Favre-Nicolin, E. Strain-Induced Magnetic Anisotropy in Epitaxial ManganiteFilms. Appl. Surf. Sci. 2002, 188, 170–175.

47. Millis, A. J.; Darling, T.; Migliori, A. Quantifying StrainDependence in “Colossal” Magnetoresistance Manga-nites. J. Appl. Phys. 1998, 83, 1588.

48. Cao, J.; Wu, J. Strain Effects in Low-Dimensional TransitionMetal Oxides. Mater. Sci. Eng., R 2011, 71, 35–52.

49. Kanki, T.; Tanaka, H.; Kawai, T. Anomalous Strain Effect inLa0.8Ba0.2MnO3 Epitaxial Thin Film: Role of the OrbitalDegree of Freedom in Stabilizing Ferromagnetism. Phys.Rev. B 2001, 64, 224418.

50. Hytch, M.; Snoeck, E.; Kilaas, R. Quantitative Measurementof Displacement and Strain Fields From HREM Micro-graphs. Ultramicroscopy 1998, 74, 131–146.

51. Hytch, M.; Houdellier, F. Mapping Stress and Strain inNanostructures by High-Resolution Transmission ElectronMicroscopy. Microelectron. Eng. 2007, 84, 460–463.

52. Yang, F.; Kemik, N.; Biegalski, M. D.; Christen, H.M.; Arenholz,E.; Takamura, Y. Strain Engineering to Control the Magneticand Magnetotransport Properties of La0.67Sr0.33MnO3 ThinFilms. Appl. Phys. Lett. 2010, 97, 092503.

53. Zubko, P.; Catalan, G.; Buckley, A.; Welche, P.; Scott, J.Strain-Gradient-Induced Polarization in SrTiO3 Single Crys-tals. Phys. Rev. Lett. 2007, 99, 167601.

54. Eliseev, E. A.; Glinchuk, M. D.; Khist, V.; Skorokhod, V. V.;Blinc, R.; Morozovska, A. N. Linear Magnetoelectric Cou-pling and Ferroelectricity Induced by the FlexomagneticEffect in Ferroics. Phys. Rev. B 2011, 84, 174112.

55. Zubko, P.; Catalan, G.; Tagantsev, A. K. Flexoelectric Effectin Solids. Annu. Rev. Mater. Res. 2013, 43, 387–421.

56. Park, S.; Lee, Y.; Prokhorov, V. Effects of Lattice-Strain-InducedDistortion and Jahn-Teller Coupling in La0.8Sr0.2MnO3/La0.8Ca0.2MnO3 Epitaxial Films. J. Korean Phys. Soc. 2004,45, 47–50.

57. Kübler, J.; William, A.; Sommers, C. Formation and Cou-pling of Magnetic Moments in Heusler Alloys. Phys. Rev. B1983, 28, 1745–1755.

58. Lampis, N.; Franchini, C.; Satta, G.; Geddo-Lehmann, A.;Massidda, S. Electronic Structure of PbFe1/2Ta1/2O3: Crys-tallographic Ordering and Magnetic Properties. Phys. Rev.B 2004, 69, 064412.

59. Cammarata, A.; Rondinelli, J. M. Octahedral Engineering ofOrbital Polarizations in Charge Transfer Oxides. Phys. Rev.B 2013, 87, 155135.

60. Maurice, J.-L. L.; Pailloux, F.; Barthélémy, A.; Durand, O.;Imhoff, D.; Lyonnet, R.; Rocher, A.; Contour, J.-P. P.; Barthel-emy, A. Strain Relaxation in the Epitaxy of La2/3Sr1/3MnO3

Grown by Pulsed-Laser Deposition on SrTiO3(001). Philos.Mag. 2003, 83, 3201–3224.

ARTIC

LE

SPURGEON ET AL . VOL. 8 ’ NO. 1 ’ 894–903 ’ 2014

www.acsnano.org

903

61. Dey, P.; Nath, T. K.; Taraphder, A. Effect of Substrate-Induced Strain on Transport and Magnetic Properties ofEpitaxial La0.67Sr0.33MnO3 Thin Films. Appl. Phys. Lett.2007, 91, 012511.

62. Haghiri-Gosnet, A.-M. A.; Renard, J.-P. CMR Manganites:Physics, Thin Films andDevices. J. Phys. D: Appl. Phys.: Appl.Phys. 2003, 36, R127–R150.

63. Pesquera, D.; Herranz, G.; Barla, A.; Pellegrin, E.; Bondino, F.;Magnano, E.; Sánchez, F.; Fontcuberta, J. Surface Symme-try-Breaking and Strain Effects on Orbital Occupancy inTransition Metal Perovskite Epitaxial Films. Nat. Commun.2012, 3, 1189.

64. Varela, M.; Lupini, A.; Benthem, K. V.; Borisevich, A.;Chisholm, M.; Shibata, N.; Abe, E.; Pennycook, S. MaterialsCharacterization in the Aberration-Corrected ScanningTransmission Electron Microscope. Annu. Rev. Mater. Res.2005, 35, 539–569.

65. Varela, M.; Oxley, M.; Luo, W.; Tao, J.; Watanabe, M.; Lupini,A.; Pantelides, S.; Pennycook, S. Atomic-Resolution Ima-ging of Oxidation States in Manganites. Phys. Rev. B 2009,79, 085117.

66. Loomer, D. B.; Al, T. A.; Weaver, L.; Cogswell, S. ManganeseValence Imaging in Mn Minerals at the Nanoscale UsingSTEM-EELS. Am. Mineral. 2007, 92, 72–79.

67. Shah, A.; Zhai, X.; Jiang, B.; Wen, J.-G.; Eckstein, J.; Zuo, J.-M.Electron Energy-Loss Study of the Electronic Structure ofAtomic Scale SrTiO3-SrMnO3-LaMnO3 Superlattices. Phys.Rev. B 2008, 77, 2–7.

68. Shah, A. B.; Ramasse, Q. M.; Zhai, X.; Wen, J. G.; May, S. J.;Petrov, I.; Bhattacharya, A.; Abbamonte, P.; Eckstein, J. N.;Zuo, J.-M. Probing Interfacial Electronic Structures inAtomic Layer LaMnO3 and SrTiO3 Superlattices. Adv.Mater. 2010, 22, 1156–1160.

69. Shah, A.; Ramasse, Q.; May, S.; Kavich, J.; Wen, J.; Zhai, X.;Eckstein, J.; Freeland, J.; Bhattacharya, A.; Zuo, J. Presenceand Spatial Distribution of Interfacial Electronic States inLaMnO3-SrMnO3 Superlattices. Phys. Rev. B 2010, 82, 1–10.

70. Schmid, H. K.; Mader, W. Oxidation States of Mn and Fe inVariousCompoundOxide Systems.Micron2006, 37, 426–432.

71. Samet, L.; Imhoff, D.; Maurice, J.-L.; Contour, J.-P.; Gloter, A.;Manoubi, T.; Fert, A.; Colliex, C. EELS Study of Interfaces inMagnetoresistive LSMO/STO/LSMO Tunnel Junctions. Eur.Phys. J. B 2003, 34, 179–192.

72. Hong, X.; Posadas, A.; Ahn, C. H. Examining the ScreeningLimit of Field Effect Devices via the Metal-Insulator Transi-tion. Appl. Phys. Lett. 2005, 86, 142501.

73. Dzero, M.; P, G. L.; Kresin, V. Z.; Gor'Kov, L. P. On Magne-toconductivity of Metallic Manganite Phases and Hetero-structures. Int. J. Mod. Phys. B 2003, 17, 2095–2115.

74. Ahn, C. H.; Di Ventra, M.; Eckstein, J. N.; Frisbie, C. D.;Gershenson, M. E.; Goldman, A. M.; Inoue, I. H.; Mannhart,J.; Millis, A. J.; Morpurgo, A. F.; et al. Electrostatic Modifica-tion of Novel Materials. Rev. Mod. Phys. 2006, 78, 1185–1212.

75. Krivanek, O. L.; Corbin, G. J.; Dellby, N.; Elston, B. F.; Keyse,R. J.; Murfitt, M. F.; Own, C. S.; Szilagyi, Z. S.; Woodruff, J. W.An Electron Microscope for the Aberration-Corrected Era.Ultramicroscopy 2008, 108, 179–95.

76. Mitchell, D. R. G.; Schaffer, B. Scripting-Customized Micro-scopy Tools for Digital Micrograph. Ultramicroscopy 2005,103, 319–332.

77. Koch, C. Determination of Core Structure Periodicity andPoint Defect Density Along Dislocations. Ph.D. thesis,Arizona State University, 2002.

78. Saito, M.; Kimoto, K.; Nagai, T.; Fukushima, S.; Akahoshi, D.;Kuwahara, H.; Matsui, Y.; Ishizuka, K. Local Crystal StructureAnalysis with 10-pm Accuracy Using Scanning Transmis-sion Electron Microscopy. J. Electron Microsc. 2009, 58,131–136.

79. Johnson, C.; Bording, J.; Zhu, Y. Structural Inhomogeneityand Twinning in YBa2Cu3O7�δ Superconductors: High-Resolution Transmission Electron Microscopy Measure-ments. Phys. Rev. B 2008, 78, 014517.

80. Dudarev, S. L.; Savrasov, S. Y.; Humphreys, C. J.; Sutton, A. P.Electron-Energy-Loss Spectra and the Structural Stability

of Nickel Oxide: An LSDAþU Study. Phys. Rev. B 1998, 57,1505–1509.

81. Rappe, A.; Rabe, K.; Kaxiras, E.; Joannopoulos, J. OptimizedPseudopotentials. Phys. Rev. B 1990, 41, 1227–1230.

82. Perdew, J. P.; Burke, K.; Ernzerhof, M. Generalized GradientApproximation Made Simple. Phys. Rev. Lett. 1996, 77,3865–3868.

83. Perdew, J.; Ruzsinszky, A.; Csonka, G.; Vydrov, O.; Scuseria,G.; Constantin, L.; Zhou, X.; Burke, K. Restoring the Density-Gradient Expansion for Exchange in Solids and Surfaces.Phys. Rev. Lett. 2008, 100, 136406.

84. Marzari, N.; Vanderbilt, D.; De Vita, A.; Payne, M. ThermalContraction and Disordering of the Al(110) Surface. Phys.Rev. Lett. 1999, 82, 3296–3299.

85. Monkhorst, H. J.; Pack, J. D. Special Points for Brillouin-Zone Integrations. Phys. Rev. B 1976, 13, 5188–5192.

86. Ma, C.; Yang, Z.; Picozzi, S.Ab Initio Electronic andMagneticStructure in La0.67Sr0.33MnO3: Strain and Correlation Ef-fects. J. Phys.: Condens. Matter 2006, 18, 7717–7728.

ARTIC

LE